A precise method of measuring 228Ra/226Ra for 226Ra determination used in U-series disequilibrium studies has been developed using the total integration method with simultaneous 228Th isobaric interference correction by multicollector ICP-MS with multiple ion counters (MIC). A sample solution of 0.2 ml was separately taken into a test tube, and all 226Ra+ and 228Ra+ signals as well as 229Th+ and 228Th16O+ from beginning to end were integrated simultaneously using four channeltrons, IC1, IC3, IC4 and IC5, respectively. Gains between IC1 and IC3 and the mass discrimination factor for 228Ra/226Ra determination were canceled out by measurement of a Ra spike solution with an accurately known Ra isotope ratio determined by TIMS, with which samples were bracketed. The isobaric interferences of 228Th+ were simultaneously corrected from the signal of 228Th16O+ and the oxide forming ratio of 228Th, which was separately determined by simultaneous measurements of 229Th+ and 229Th16O+ by IC3 and IC5. The gains between IC3 and IC5 for the Th correction were also canceled out. The ion yield of Ra in MIC-ICP-MS was ∼1.5%. The intermediate precision using 6.6, 3.3, 1.7, 0.83 and 0.41 fg of 226Ra with 228Ra/226Ra = ∼5 was 0.33, 0.93, 0.76, 1.6 and 1.5%, respectively. The 228Ra/226Ra ratio showed no systematic change dependent on 228Th/228Ra up to 0.25, indicating the validity of the 228Th correction. At similar amounts of Ra, our method gives intermediate precision values similar to or better than previous MC-ICP-MS and TIMS studies. The precision of the method was verified using another Ra spike with 228Ra/226Ra = ∼1 determined by TIMS. The analytical performance of the method was further investigated using the silicate samples JB-2 and JB-3 basalts issued by GSJ.